Face Generation¶

In this project, you'll use generative adversarial networks to generate new images of faces.

Get the Data¶

You'll be using two datasets in this project:

MNIST
CelebA

Since the celebA dataset is complex and you're doing GANs in a project for the first time, we want you to test your neural network on MNIST before CelebA. Running the GANs on MNIST will allow you to see how well your model trains sooner.

If you're using FloydHub, set data_dir to "/input" and use the FloydHub data ID "R5KrjnANiKVhLWAkpXhNBe".

data_dir = './data'

# FloydHub - Use with data ID "R5KrjnANiKVhLWAkpXhNBe"
#data_dir = '/input'


"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
import helper

helper.download_extract('mnist', data_dir)
helper.download_extract('celeba', data_dir)

Found mnist Data
Found celeba Data

alpha = 0.15

Explore the Data¶

MNIST¶

As you're aware, the MNIST dataset contains images of handwritten digits. You can view the first number of examples by changing show_n_images.

show_n_images = 25

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
%matplotlib inline
import os
from glob import glob
from matplotlib import pyplot

mnist_images = helper.get_batch(glob(os.path.join(data_dir, 'mnist/*.jpg'))[:show_n_images], 28, 28, 'L')
pyplot.imshow(helper.images_square_grid(mnist_images, 'L'), cmap='gray')

<matplotlib.image.AxesImage at 0x7fce36ed61d0>

CelebA¶

The CelebFaces Attributes Dataset (CelebA) dataset contains over 200,000 celebrity images with annotations. Since you're going to be generating faces, you won't need the annotations. You can view the first number of examples by changing show_n_images.

show_n_images = 25

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
mnist_images = helper.get_batch(glob(os.path.join(data_dir, 'img_align_celeba/*.jpg'))[:show_n_images], 28, 28, 'RGB')
pyplot.imshow(helper.images_square_grid(mnist_images, 'RGB'))

<matplotlib.image.AxesImage at 0x7fce36e13c50>

Preprocess the Data¶

Since the project's main focus is on building the GANs, we'll preprocess the data for you. The values of the MNIST and CelebA dataset will be in the range of -0.5 to 0.5 of 28x28 dimensional images. The CelebA images will be cropped to remove parts of the image that don't include a face, then resized down to 28x28.

The MNIST images are black and white images with a single color channel while the CelebA images have 3 color channels (RGB color channel).

Build the Neural Network¶

You'll build the components necessary to build a GANs by implementing the following functions below:

model_inputs
discriminator
generator
model_loss
model_opt
train

Check the Version of TensorFlow and Access to GPU¶

This will check to make sure you have the correct version of TensorFlow and access to a GPU

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
from distutils.version import LooseVersion
import warnings
import tensorflow as tf

# Check TensorFlow Version
assert LooseVersion(tf.__version__) >= LooseVersion('1.0'), 'Please use TensorFlow version 1.0 or newer.  You are using {}'.format(tf.__version__)
print('TensorFlow Version: {}'.format(tf.__version__))

# Check for a GPU
if not tf.test.gpu_device_name():
    warnings.warn('No GPU found. Please use a GPU to train your neural network.')
else:
    print('Default GPU Device: {}'.format(tf.test.gpu_device_name()))

TensorFlow Version: 1.0.1
Default GPU Device: /gpu:0

Input¶

Implement the model_inputs function to create TF Placeholders for the Neural Network. It should create the following placeholders:

Real input images placeholder with rank 4 using image_width, image_height, and image_channels.
Z input placeholder with rank 2 using z_dim.
Learning rate placeholder with rank 0.

Return the placeholders in the following the tuple (tensor of real input images, tensor of z data)

import problem_unittests as tests

def model_inputs(image_width, image_height, image_channels, z_dim):
    """
    Create the model inputs
    :param image_width: The input image width
    :param image_height: The input image height
    :param image_channels: The number of image channels
    :param z_dim: The dimension of Z
    :return: Tuple of (tensor of real input images, tensor of z data, learning rate)
    """
    # TODO: Implement Function
    img = tf.placeholder(tf.float32, shape=[None, image_width, image_height, image_channels], name='input_z')
    inputs_z = tf.placeholder(tf.float32, shape=[None, z_dim], name='input_z')
    lr = tf.placeholder(tf.float32, name='lr')
    return img, inputs_z, lr


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_inputs(model_inputs)

Tests Passed

Discriminator¶

Implement discriminator to create a discriminator neural network that discriminates on images. This function should be able to reuse the variabes in the neural network. Use tf.variable_scope with a scope name of "discriminator" to allow the variables to be reused. The function should return a tuple of (tensor output of the generator, tensor logits of the generator).

def discriminator(images, reuse=False):
    """
    Create the discriminator network
    :param image: Tensor of input image(s)
    :param reuse: Boolean if the weights should be reused
    :return: Tuple of (tensor output of the discriminator, tensor logits of the discriminator)
    """
    
#     print('discriminator')
#     print('images: ', images)
    # TODO: Implement Function
    with tf.variable_scope('discriminator', reuse=reuse):
        # Input layer is ?x28x28x3
        x1 = tf.layers.conv2d(images, 64, 5, strides=2, padding='same', 
                              kernel_initializer = tf.random_normal_initializer(stddev=0.02))
        relu1 = tf.maximum(0.02 * x1, x1)
        relu1 = tf.layers.dropout(relu1, rate=0.5)
        # 14x14x64
#         print(x1)
        x2 = tf.layers.conv2d(relu1, 128, 5, strides=2, padding='same', 
                              kernel_initializer = tf.random_normal_initializer(stddev=0.02))
        bn2 = tf.layers.batch_normalization(x2, training=True)
        relu2 = tf.maximum(0.02 * bn2, bn2)
        relu2 = tf.layers.dropout(relu2, rate=0.5)
        # 7x7x128
#         print(x2)
        x3 = tf.layers.conv2d(relu2, 256, 5, strides=2, padding='same', 
                              kernel_initializer = tf.random_normal_initializer(stddev=0.02))
        bn3 = tf.layers.batch_normalization(x3, training=True)
        relu3 = tf.maximum(0.02 * bn3, bn3)
        relu3 = tf.layers.dropout(relu3, rate=0.5)
        # 4x4x256 
#         print(x3)
        # Flatten it
        flat = tf.reshape(relu3, (-1, 4*4*256))
        logits = tf.layers.dense(flat, 1)
#         print(logits)
        out = tf.sigmoid(logits)
#         print('discriminator out: ', out)
        return out, logits



"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_discriminator(discriminator, tf)

Tests Passed

Generator¶

Implement generator to generate an image using z. This function should be able to reuse the variabes in the neural network. Use tf.variable_scope with a scope name of "generator" to allow the variables to be reused. The function should return the generated 28 x 28 x out_channel_dim images.

def generator(z, out_channel_dim, is_train=True):
    """
    Create the generator network
    :param z: Input z
    :param out_channel_dim: The number of channels in the output image
    :param is_train: Boolean if generator is being used for training
    :return: The tensor output of the generator
    """
#     print('generator')
#     print('out_channel_dim: ', out_channel_dim)
    # TODO: Implement Function
    with tf.variable_scope('generator', reuse=not is_train):
        # First fully connected layer
        x1 = tf.layers.dense(z, 7*7*512)
        # Reshape it to start the convolutional stack
        x1 = tf.reshape(x1, (-1, 7, 7, 512))
#         x1 = tf.layers.batch_normalization(x1, training=training)
        x1 = tf.maximum(alpha * x1, x1)
        # 7x7x512 now
#         print(x1)
        x2 = tf.layers.conv2d_transpose(x1, 256, 5, strides=1, padding='same')
        x2 = tf.layers.batch_normalization(x2, training=is_train)
        x2 = tf.maximum(alpha * x2, x2)
        # 7x7x256 now
#         print(x2)
        x3 = tf.layers.conv2d_transpose(x2, 128, 5, strides=2, padding='same')
        x3 = tf.layers.batch_normalization(x3, training=is_train)
        x3 = tf.maximum(alpha * x3, x3)
        # 14x14x128 now
#         print(x3)

        x4 = tf.layers.conv2d_transpose(x3, 64, 5, strides=2, padding='same')
        x4 = tf.layers.batch_normalization(x4, training=is_train)
        x4 = tf.maximum(alpha * x4, x4)
        
        
        # Output layer
        logits = tf.layers.conv2d_transpose(x4, out_channel_dim, 5, strides=1, padding='same')
        # 28x28x3 now
#         print(logits)3
        out = tf.tanh(logits)
        
        return out


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_generator(generator, tf)

Tests Passed

Loss¶

Implement model_loss to build the GANs for training and calculate the loss. The function should return a tuple of (discriminator loss, generator loss). Use the following functions you implemented:

discriminator(images, reuse=False)
generator(z, out_channel_dim, is_train=True)

def model_loss(input_real, input_z, out_channel_dim):
    """
    Get the loss for the discriminator and generator
    :param input_real: Images from the real dataset
    :param input_z: Z input
    :param out_channel_dim: The number of channels in the output image
    :return: A tuple of (discriminator loss, generator loss)
    """
    # TODO: Implement Function
    
#     print('Generator for fake images...')
    g_model = generator(input_z, out_channel_dim)
#     print('Passing discriminator with real images...')
    d_model_real, d_logits_real = discriminator(input_real)
#     print('Passing discriminator with fake images...')
    d_model_fake, d_logits_fake = discriminator(g_model, reuse=True)

    d_loss_real = tf.reduce_mean(
        tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_real, labels=tf.ones_like(d_model_real)))
    d_loss_fake = tf.reduce_mean(
        tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_fake, labels=tf.zeros_like(d_model_fake)))
    g_loss = tf.reduce_mean(
        tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_fake, labels=tf.ones_like(d_model_fake)))

    d_loss = d_loss_real + d_loss_fake

    return d_loss, g_loss


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_loss(model_loss)

Tests Passed

Optimization¶

Implement model_opt to create the optimization operations for the GANs. Use tf.trainable_variables to get all the trainable variables. Filter the variables with names that are in the discriminator and generator scope names. The function should return a tuple of (discriminator training operation, generator training operation).

def model_opt(d_loss, g_loss, learning_rate, beta1):
    """
    Get optimization operations
    :param d_loss: Discriminator loss Tensor
    :param g_loss: Generator loss Tensor
    :param learning_rate: Learning Rate Placeholder
    :param beta1: The exponential decay rate for the 1st moment in the optimizer
    :return: A tuple of (discriminator training operation, generator training operation)
    """
    # TODO: Implement Function
    
    # Get weights and bias to update
    t_vars = tf.trainable_variables()
    d_vars = [var for var in t_vars if var.name.startswith('discriminator')]
    g_vars = [var for var in t_vars if var.name.startswith('generator')]

    # Optimize
    d_train_opt = tf.train.AdamOptimizer(learning_rate, beta1=beta1).minimize(d_loss, var_list=d_vars)
    
    update_ops = tf.get_collection(tf.GraphKeys.UPDATE_OPS)
    g_updates = [opt for opt in update_ops if opt.name.startswith('generator')]
    with tf.control_dependencies(g_updates):
        g_train_opt = tf.train.AdamOptimizer(learning_rate, beta1=beta1).minimize(g_loss, var_list=g_vars)

    return d_train_opt, g_train_opt


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_opt(model_opt, tf)

Tests Passed

Neural Network Training¶

Show Output¶

Use this function to show the current output of the generator during training. It will help you determine how well the GANs is training.

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
import numpy as np

def show_generator_output(sess, n_images, input_z, out_channel_dim, image_mode):
    """
    Show example output for the generator
    :param sess: TensorFlow session
    :param n_images: Number of Images to display
    :param input_z: Input Z Tensor
    :param out_channel_dim: The number of channels in the output image
    :param image_mode: The mode to use for images ("RGB" or "L")
    """
    cmap = None if image_mode == 'RGB' else 'gray'
    z_dim = input_z.get_shape().as_list()[-1]
    example_z = np.random.uniform(-1, 1, size=[n_images, z_dim])

    samples = sess.run(
        generator(input_z, out_channel_dim, False),
        feed_dict={input_z: example_z})

    images_grid = helper.images_square_grid(samples, image_mode)
    pyplot.imshow(images_grid, cmap=cmap)
    pyplot.show()

Train¶

Implement train to build and train the GANs. Use the following functions you implemented:

model_inputs(image_width, image_height, image_channels, z_dim)
model_loss(input_real, input_z, out_channel_dim)
model_opt(d_loss, g_loss, learning_rate, beta1)

Use the show_generator_output to show generator output while you train. Running show_generator_output for every batch will drastically increase training time and increase the size of the notebook. It's recommended to print the generator output every 100 batches.

# tf.reset_default_graph()
show_every=20
print_every=1

def train(epoch_count, batch_size, z_dim, learning_rate, beta1, get_batches, data_shape, data_image_mode):
    """
    Train the GAN
    :param epoch_count: Number of epochs
    :param batch_size: Batch Size
    :param z_dim: Z dimension
    :param learning_rate: Learning Rate
    :param beta1: The exponential decay rate for the 1st moment in the optimizer
    :param get_batches: Function to get batches
    :param data_shape: Shape of the data
    :param data_image_mode: The image mode to use for images ("RGB" or "L")
    """
    # TODO: Build Model
    
    print(epoch_count)
    print(batch_size)
    print(z_dim)
    print(learning_rate)
    print(beta1)
    print(data_shape)
    print(data_image_mode)
    
    

    image_width = data_shape[1]
    image_height = data_shape[2]
    image_channels = data_shape[3]
    input_real, input_z, learning_rate_ = model_inputs(image_width, image_height, image_channels, z_dim)
    d_loss, g_loss = model_loss(input_real, input_z, image_channels)
    d_train_opt, g_train_opt = model_opt(d_loss, g_loss, learning_rate, beta1)
    
    steps = 0
    print('Starting Session...')
    with tf.Session() as sess:
        sess.run(tf.global_variables_initializer())
        for epoch_i in range(epoch_count):
            print('Epoch :', epoch_i)
            for batch_images in get_batches(batch_size):
                # TODO: Train Model
                steps +=1
                
                batch_images = batch_images*2.0
                
                batch_z = np.random.uniform(-1, 1, size=(batch_size, z_dim))

                _ = sess.run(d_train_opt, feed_dict={input_real: batch_images, input_z: batch_z})
                _ = sess.run(g_train_opt, feed_dict={input_z: batch_z})

                
                if steps % print_every == 0:
                    # At the end of each epoch, get the losses and print them out
                    train_loss_d = d_loss.eval({input_z: batch_z, input_real: batch_images})
                    train_loss_g = g_loss.eval({input_z: batch_z})

                    print("Epoch {}/{}...".format(epoch_i+1, epoch_count),
                          "Discriminator Loss: {:.4f}...".format(train_loss_d),
                          "Generator Loss: {:.4f}".format(train_loss_g))
                    
                    
                    
                if steps % show_every == 0:
                    show_generator_output(sess, 16, input_z, image_channels, data_image_mode)

MNIST¶

Test your GANs architecture on MNIST. After 2 epochs, the GANs should be able to generate images that look like handwritten digits. Make sure the loss of the generator is lower than the loss of the discriminator or close to 0.

batch_size = 64
z_dim = 30
learning_rate = 0.001
beta1 = 0.4
tf.reset_default_graph()
"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
epochs = 2

mnist_dataset = helper.Dataset('mnist', glob(os.path.join(data_dir, 'mnist/*.jpg')))
with tf.Graph().as_default():
    train(epochs, batch_size, z_dim, learning_rate, beta1, mnist_dataset.get_batches,
          mnist_dataset.shape, mnist_dataset.image_mode)

2
64
30
0.001
0.4
(60000, 28, 28, 1)
L
Starting Session...
Epoch : 0
Epoch 1/2... Discriminator Loss: 15.6208... Generator Loss: 0.0000
Epoch 1/2... Discriminator Loss: 9.0676... Generator Loss: 0.0002
Epoch 1/2... Discriminator Loss: 3.1156... Generator Loss: 3.0856
Epoch 1/2... Discriminator Loss: 8.0054... Generator Loss: 0.0005
Epoch 1/2... Discriminator Loss: 0.7265... Generator Loss: 1.3715
Epoch 1/2... Discriminator Loss: 1.0292... Generator Loss: 1.2111
Epoch 1/2... Discriminator Loss: 0.5375... Generator Loss: 3.9218
Epoch 1/2... Discriminator Loss: 3.2712... Generator Loss: 0.1677
Epoch 1/2... Discriminator Loss: 2.1505... Generator Loss: 9.4128
Epoch 1/2... Discriminator Loss: 0.3546... Generator Loss: 2.6493
Epoch 1/2... Discriminator Loss: 0.4817... Generator Loss: 1.7825
Epoch 1/2... Discriminator Loss: 1.5662... Generator Loss: 0.7328
Epoch 1/2... Discriminator Loss: 2.3180... Generator Loss: 5.8899
Epoch 1/2... Discriminator Loss: 2.3334... Generator Loss: 0.8717
Epoch 1/2... Discriminator Loss: 0.4765... Generator Loss: 2.0655
Epoch 1/2... Discriminator Loss: 0.7843... Generator Loss: 2.6188
Epoch 1/2... Discriminator Loss: 2.8057... Generator Loss: 0.0851
Epoch 1/2... Discriminator Loss: 3.9585... Generator Loss: 7.0877
Epoch 1/2... Discriminator Loss: 1.0285... Generator Loss: 1.0105
Epoch 1/2... Discriminator Loss: 1.0583... Generator Loss: 0.8424

Epoch 1/2... Discriminator Loss: 0.6501... Generator Loss: 1.8380
Epoch 1/2... Discriminator Loss: 1.2884... Generator Loss: 0.7988
Epoch 1/2... Discriminator Loss: 0.9662... Generator Loss: 2.1217
Epoch 1/2... Discriminator Loss: 1.7815... Generator Loss: 0.4475
Epoch 1/2... Discriminator Loss: 1.7089... Generator Loss: 1.7101
Epoch 1/2... Discriminator Loss: 2.7239... Generator Loss: 0.1279
Epoch 1/2... Discriminator Loss: 2.6157... Generator Loss: 3.5681
Epoch 1/2... Discriminator Loss: 2.2232... Generator Loss: 0.5533
Epoch 1/2... Discriminator Loss: 1.3289... Generator Loss: 1.8077
Epoch 1/2... Discriminator Loss: 1.8193... Generator Loss: 0.3076
Epoch 1/2... Discriminator Loss: 1.6719... Generator Loss: 1.5953
Epoch 1/2... Discriminator Loss: 1.3133... Generator Loss: 0.6314
Epoch 1/2... Discriminator Loss: 0.8222... Generator Loss: 1.4861
Epoch 1/2... Discriminator Loss: 1.3087... Generator Loss: 0.6197
Epoch 1/2... Discriminator Loss: 1.2389... Generator Loss: 1.6094
Epoch 1/2... Discriminator Loss: 1.6728... Generator Loss: 0.4040
Epoch 1/2... Discriminator Loss: 1.3386... Generator Loss: 2.3284
Epoch 1/2... Discriminator Loss: 1.6889... Generator Loss: 0.4054
Epoch 1/2... Discriminator Loss: 1.2259... Generator Loss: 1.4781
Epoch 1/2... Discriminator Loss: 1.4976... Generator Loss: 0.5101

Epoch 1/2... Discriminator Loss: 1.3207... Generator Loss: 1.2700
Epoch 1/2... Discriminator Loss: 1.4415... Generator Loss: 0.5087
Epoch 1/2... Discriminator Loss: 1.5859... Generator Loss: 2.4874
Epoch 1/2... Discriminator Loss: 1.9187... Generator Loss: 0.2208
Epoch 1/2... Discriminator Loss: 1.8620... Generator Loss: 2.8884
Epoch 1/2... Discriminator Loss: 1.7599... Generator Loss: 0.3773
Epoch 1/2... Discriminator Loss: 1.1094... Generator Loss: 1.5529
Epoch 1/2... Discriminator Loss: 1.0405... Generator Loss: 0.9657
Epoch 1/2... Discriminator Loss: 0.9580... Generator Loss: 1.2174
Epoch 1/2... Discriminator Loss: 1.0976... Generator Loss: 0.7437
Epoch 1/2... Discriminator Loss: 0.8341... Generator Loss: 1.4085
Epoch 1/2... Discriminator Loss: 0.8064... Generator Loss: 1.0416
Epoch 1/2... Discriminator Loss: 1.0010... Generator Loss: 0.8551
Epoch 1/2... Discriminator Loss: 1.3553... Generator Loss: 0.8101
Epoch 1/2... Discriminator Loss: 1.3556... Generator Loss: 1.1402
Epoch 1/2... Discriminator Loss: 1.1698... Generator Loss: 0.5891
Epoch 1/2... Discriminator Loss: 1.1795... Generator Loss: 2.5278
Epoch 1/2... Discriminator Loss: 2.2399... Generator Loss: 0.1809
Epoch 1/2... Discriminator Loss: 1.8066... Generator Loss: 2.2590
Epoch 1/2... Discriminator Loss: 1.6818... Generator Loss: 0.3519

Epoch 1/2... Discriminator Loss: 1.2962... Generator Loss: 1.3949
Epoch 1/2... Discriminator Loss: 1.3425... Generator Loss: 0.4806
Epoch 1/2... Discriminator Loss: 1.0882... Generator Loss: 1.4747
Epoch 1/2... Discriminator Loss: 1.1613... Generator Loss: 0.6159
Epoch 1/2... Discriminator Loss: 1.0978... Generator Loss: 1.5265
Epoch 1/2... Discriminator Loss: 1.1069... Generator Loss: 0.6818
Epoch 1/2... Discriminator Loss: 1.1626... Generator Loss: 1.0649
Epoch 1/2... Discriminator Loss: 1.2750... Generator Loss: 0.6878
Epoch 1/2... Discriminator Loss: 1.1819... Generator Loss: 0.8401
Epoch 1/2... Discriminator Loss: 1.2076... Generator Loss: 0.9922
Epoch 1/2... Discriminator Loss: 1.3328... Generator Loss: 0.5451
Epoch 1/2... Discriminator Loss: 1.1909... Generator Loss: 1.7594
Epoch 1/2... Discriminator Loss: 1.4927... Generator Loss: 0.3705
Epoch 1/2... Discriminator Loss: 1.4588... Generator Loss: 2.0271
Epoch 1/2... Discriminator Loss: 1.8541... Generator Loss: 0.3059
Epoch 1/2... Discriminator Loss: 1.7747... Generator Loss: 2.7908
Epoch 1/2... Discriminator Loss: 1.5673... Generator Loss: 0.3894
Epoch 1/2... Discriminator Loss: 1.0191... Generator Loss: 1.6999
Epoch 1/2... Discriminator Loss: 1.3484... Generator Loss: 0.5575
Epoch 1/2... Discriminator Loss: 1.3561... Generator Loss: 1.0387

Epoch 1/2... Discriminator Loss: 1.3869... Generator Loss: 0.5087
Epoch 1/2... Discriminator Loss: 1.3363... Generator Loss: 1.2697
Epoch 1/2... Discriminator Loss: 1.6046... Generator Loss: 0.4209
Epoch 1/2... Discriminator Loss: 1.4897... Generator Loss: 1.1738
Epoch 1/2... Discriminator Loss: 1.4858... Generator Loss: 0.4308
Epoch 1/2... Discriminator Loss: 1.2366... Generator Loss: 1.5633
Epoch 1/2... Discriminator Loss: 1.4473... Generator Loss: 0.4318
Epoch 1/2... Discriminator Loss: 1.4423... Generator Loss: 1.7019
Epoch 1/2... Discriminator Loss: 1.2075... Generator Loss: 0.5340
Epoch 1/2... Discriminator Loss: 1.0974... Generator Loss: 1.2940
Epoch 1/2... Discriminator Loss: 1.2261... Generator Loss: 0.5685
Epoch 1/2... Discriminator Loss: 1.0757... Generator Loss: 1.4096
Epoch 1/2... Discriminator Loss: 1.3372... Generator Loss: 0.5716
Epoch 1/2... Discriminator Loss: 1.3172... Generator Loss: 1.7041
Epoch 1/2... Discriminator Loss: 1.7872... Generator Loss: 0.2998
Epoch 1/2... Discriminator Loss: 1.5822... Generator Loss: 1.8892
Epoch 1/2... Discriminator Loss: 1.1291... Generator Loss: 0.7050
Epoch 1/2... Discriminator Loss: 1.0876... Generator Loss: 0.8118
Epoch 1/2... Discriminator Loss: 1.1626... Generator Loss: 0.9160
Epoch 1/2... Discriminator Loss: 1.1490... Generator Loss: 1.0097

Epoch 1/2... Discriminator Loss: 1.2158... Generator Loss: 0.6923
Epoch 1/2... Discriminator Loss: 1.1501... Generator Loss: 0.9800
Epoch 1/2... Discriminator Loss: 1.3536... Generator Loss: 0.5056
Epoch 1/2... Discriminator Loss: 1.6712... Generator Loss: 1.6707
Epoch 1/2... Discriminator Loss: 1.5974... Generator Loss: 0.3244
Epoch 1/2... Discriminator Loss: 1.3684... Generator Loss: 1.6633
Epoch 1/2... Discriminator Loss: 1.3283... Generator Loss: 0.5004
Epoch 1/2... Discriminator Loss: 1.2336... Generator Loss: 0.9394
Epoch 1/2... Discriminator Loss: 1.3159... Generator Loss: 0.6315
Epoch 1/2... Discriminator Loss: 1.3660... Generator Loss: 0.8817
Epoch 1/2... Discriminator Loss: 1.3895... Generator Loss: 0.5088
Epoch 1/2... Discriminator Loss: 1.4992... Generator Loss: 0.8962
Epoch 1/2... Discriminator Loss: 1.3986... Generator Loss: 0.6248
Epoch 1/2... Discriminator Loss: 1.1684... Generator Loss: 1.0472
Epoch 1/2... Discriminator Loss: 1.2179... Generator Loss: 0.6599
Epoch 1/2... Discriminator Loss: 1.2714... Generator Loss: 1.0560
Epoch 1/2... Discriminator Loss: 1.5823... Generator Loss: 0.4099
Epoch 1/2... Discriminator Loss: 1.6600... Generator Loss: 1.4811
Epoch 1/2... Discriminator Loss: 1.7359... Generator Loss: 0.2980
Epoch 1/2... Discriminator Loss: 1.5236... Generator Loss: 1.4038

Epoch 1/2... Discriminator Loss: 1.6516... Generator Loss: 0.3536
Epoch 1/2... Discriminator Loss: 1.3528... Generator Loss: 1.3009
Epoch 1/2... Discriminator Loss: 1.3335... Generator Loss: 0.6669
Epoch 1/2... Discriminator Loss: 1.3385... Generator Loss: 0.8412
Epoch 1/2... Discriminator Loss: 1.1262... Generator Loss: 0.8262
Epoch 1/2... Discriminator Loss: 1.0767... Generator Loss: 0.8213
Epoch 1/2... Discriminator Loss: 1.2093... Generator Loss: 0.9352
Epoch 1/2... Discriminator Loss: 1.3266... Generator Loss: 0.5761
Epoch 1/2... Discriminator Loss: 1.4833... Generator Loss: 0.9769
Epoch 1/2... Discriminator Loss: 1.4714... Generator Loss: 0.4071
Epoch 1/2... Discriminator Loss: 1.4066... Generator Loss: 1.2835
Epoch 1/2... Discriminator Loss: 1.4300... Generator Loss: 0.4331
Epoch 1/2... Discriminator Loss: 1.2633... Generator Loss: 1.5096
Epoch 1/2... Discriminator Loss: 1.5138... Generator Loss: 0.4460
Epoch 1/2... Discriminator Loss: 1.3116... Generator Loss: 0.9760
Epoch 1/2... Discriminator Loss: 1.2716... Generator Loss: 0.7527
Epoch 1/2... Discriminator Loss: 1.2745... Generator Loss: 0.7830
Epoch 1/2... Discriminator Loss: 1.2020... Generator Loss: 0.7517
Epoch 1/2... Discriminator Loss: 1.0940... Generator Loss: 1.0240
Epoch 1/2... Discriminator Loss: 1.3267... Generator Loss: 0.5477

Epoch 1/2... Discriminator Loss: 1.3022... Generator Loss: 1.1654
Epoch 1/2... Discriminator Loss: 1.5476... Generator Loss: 0.4363
Epoch 1/2... Discriminator Loss: 1.3412... Generator Loss: 1.3678
Epoch 1/2... Discriminator Loss: 1.4216... Generator Loss: 0.4306
Epoch 1/2... Discriminator Loss: 1.3117... Generator Loss: 1.0673
Epoch 1/2... Discriminator Loss: 1.3474... Generator Loss: 0.6407
Epoch 1/2... Discriminator Loss: 1.3080... Generator Loss: 0.8790
Epoch 1/2... Discriminator Loss: 1.2809... Generator Loss: 0.6759
Epoch 1/2... Discriminator Loss: 1.2630... Generator Loss: 0.8684
Epoch 1/2... Discriminator Loss: 1.1794... Generator Loss: 0.9322
Epoch 1/2... Discriminator Loss: 1.3095... Generator Loss: 0.6387
Epoch 1/2... Discriminator Loss: 1.3741... Generator Loss: 0.9465
Epoch 1/2... Discriminator Loss: 1.3256... Generator Loss: 0.6516
Epoch 1/2... Discriminator Loss: 1.2387... Generator Loss: 0.7112
Epoch 1/2... Discriminator Loss: 1.1188... Generator Loss: 1.1177
Epoch 1/2... Discriminator Loss: 1.2297... Generator Loss: 0.5929
Epoch 1/2... Discriminator Loss: 1.3541... Generator Loss: 1.2881
Epoch 1/2... Discriminator Loss: 1.4474... Generator Loss: 0.3986
Epoch 1/2... Discriminator Loss: 1.3440... Generator Loss: 1.4614
Epoch 1/2... Discriminator Loss: 1.5437... Generator Loss: 0.4016

Epoch 1/2... Discriminator Loss: 1.4401... Generator Loss: 1.2547
Epoch 1/2... Discriminator Loss: 1.1554... Generator Loss: 0.6857
Epoch 1/2... Discriminator Loss: 1.3382... Generator Loss: 0.5922
Epoch 1/2... Discriminator Loss: 1.4117... Generator Loss: 1.0724
Epoch 1/2... Discriminator Loss: 1.2603... Generator Loss: 0.5661
Epoch 1/2... Discriminator Loss: 1.2810... Generator Loss: 1.0895
Epoch 1/2... Discriminator Loss: 1.2034... Generator Loss: 0.5905
Epoch 1/2... Discriminator Loss: 1.2074... Generator Loss: 1.1226
Epoch 1/2... Discriminator Loss: 1.2283... Generator Loss: 0.6585
Epoch 1/2... Discriminator Loss: 1.2293... Generator Loss: 0.8966
Epoch 1/2... Discriminator Loss: 1.1830... Generator Loss: 0.7730
Epoch 1/2... Discriminator Loss: 1.1398... Generator Loss: 0.9304
Epoch 1/2... Discriminator Loss: 1.2085... Generator Loss: 0.6797
Epoch 1/2... Discriminator Loss: 1.2853... Generator Loss: 0.9042
Epoch 1/2... Discriminator Loss: 1.2467... Generator Loss: 0.6230
Epoch 1/2... Discriminator Loss: 1.2643... Generator Loss: 0.9226
Epoch 1/2... Discriminator Loss: 1.2630... Generator Loss: 0.7049
Epoch 1/2... Discriminator Loss: 1.2331... Generator Loss: 0.7327
Epoch 1/2... Discriminator Loss: 1.2659... Generator Loss: 0.9665
Epoch 1/2... Discriminator Loss: 1.3337... Generator Loss: 0.5115

Epoch 1/2... Discriminator Loss: 1.3408... Generator Loss: 1.6115
Epoch 1/2... Discriminator Loss: 1.3738... Generator Loss: 0.4632
Epoch 1/2... Discriminator Loss: 1.4004... Generator Loss: 1.2416
Epoch 1/2... Discriminator Loss: 1.2220... Generator Loss: 0.7197
Epoch 1/2... Discriminator Loss: 1.2381... Generator Loss: 0.7954
Epoch 1/2... Discriminator Loss: 1.2517... Generator Loss: 0.8737
Epoch 1/2... Discriminator Loss: 1.1539... Generator Loss: 0.7178
Epoch 1/2... Discriminator Loss: 1.2352... Generator Loss: 0.7667
Epoch 1/2... Discriminator Loss: 1.1796... Generator Loss: 0.9664
Epoch 1/2... Discriminator Loss: 1.1555... Generator Loss: 0.6237
Epoch 1/2... Discriminator Loss: 1.1698... Generator Loss: 1.3071
Epoch 1/2... Discriminator Loss: 1.4290... Generator Loss: 0.4295
Epoch 1/2... Discriminator Loss: 1.2537... Generator Loss: 1.4099
Epoch 1/2... Discriminator Loss: 1.4660... Generator Loss: 0.4017
Epoch 1/2... Discriminator Loss: 1.3563... Generator Loss: 1.2847
Epoch 1/2... Discriminator Loss: 1.5706... Generator Loss: 0.3731
Epoch 1/2... Discriminator Loss: 1.3072... Generator Loss: 1.2480
Epoch 1/2... Discriminator Loss: 1.2298... Generator Loss: 0.6436
Epoch 1/2... Discriminator Loss: 1.1113... Generator Loss: 0.9661
Epoch 1/2... Discriminator Loss: 1.2748... Generator Loss: 0.6251

CelebA¶

Run your GANs on CelebA. It will take around 20 minutes on the average GPU to run one epoch. You can run the whole epoch or stop when it starts to generate realistic faces.

batch_size = 64
z_dim = 100
learning_rate = 0.001
beta1 = 0.4
tf.reset_default_graph()


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
epochs = 1

celeba_dataset = helper.Dataset('celeba', glob(os.path.join(data_dir, 'img_align_celeba/*.jpg')))
with tf.Graph().as_default():
    train(epochs, batch_size, z_dim, learning_rate, beta1, celeba_dataset.get_batches,
          celeba_dataset.shape, celeba_dataset.image_mode)

1
64
100
0.001
0.4
(202599, 28, 28, 3)
RGB
Starting Session...
Epoch : 0
Epoch 1/1... Discriminator Loss: 6.6300... Generator Loss: 0.0020
Epoch 1/1... Discriminator Loss: 2.4759... Generator Loss: 4.4734
Epoch 1/1... Discriminator Loss: 8.3581... Generator Loss: 0.0003
Epoch 1/1... Discriminator Loss: 0.4677... Generator Loss: 8.5955
Epoch 1/1... Discriminator Loss: 0.1876... Generator Loss: 3.6293
Epoch 1/1... Discriminator Loss: 3.9052... Generator Loss: 0.0279
Epoch 1/1... Discriminator Loss: 2.2565... Generator Loss: 16.6714
Epoch 1/1... Discriminator Loss: 0.1648... Generator Loss: 10.1435
Epoch 1/1... Discriminator Loss: 0.2359... Generator Loss: 1.8105
Epoch 1/1... Discriminator Loss: 1.1185... Generator Loss: 0.5585
Epoch 1/1... Discriminator Loss: 0.4542... Generator Loss: 16.3597
Epoch 1/1... Discriminator Loss: 0.5135... Generator Loss: 5.7610
Epoch 1/1... Discriminator Loss: 2.8440... Generator Loss: 0.0926
Epoch 1/1... Discriminator Loss: 1.1611... Generator Loss: 13.3102
Epoch 1/1... Discriminator Loss: 0.2302... Generator Loss: 8.0784
Epoch 1/1... Discriminator Loss: 2.8248... Generator Loss: 0.1846
Epoch 1/1... Discriminator Loss: 1.7292... Generator Loss: 15.0618
Epoch 1/1... Discriminator Loss: 0.1999... Generator Loss: 8.5550
Epoch 1/1... Discriminator Loss: 2.2119... Generator Loss: 0.3061
Epoch 1/1... Discriminator Loss: 0.6973... Generator Loss: 9.5725

Epoch 1/1... Discriminator Loss: 0.9230... Generator Loss: 2.2232
Epoch 1/1... Discriminator Loss: 1.4939... Generator Loss: 0.8583
Epoch 1/1... Discriminator Loss: 0.8319... Generator Loss: 4.5062
Epoch 1/1... Discriminator Loss: 5.7662... Generator Loss: 0.0114
Epoch 1/1... Discriminator Loss: 4.7892... Generator Loss: 13.3787
Epoch 1/1... Discriminator Loss: 0.5154... Generator Loss: 3.0066
Epoch 1/1... Discriminator Loss: 4.3471... Generator Loss: 0.0854
Epoch 1/1... Discriminator Loss: 3.7445... Generator Loss: 7.8703
Epoch 1/1... Discriminator Loss: 1.3616... Generator Loss: 1.0955
Epoch 1/1... Discriminator Loss: 2.0019... Generator Loss: 0.3913
Epoch 1/1... Discriminator Loss: 3.0854... Generator Loss: 5.5523
Epoch 1/1... Discriminator Loss: 1.3466... Generator Loss: 0.5577
Epoch 1/1... Discriminator Loss: 0.8195... Generator Loss: 1.4339
Epoch 1/1... Discriminator Loss: 0.7353... Generator Loss: 1.4591
Epoch 1/1... Discriminator Loss: 1.1121... Generator Loss: 0.6445
Epoch 1/1... Discriminator Loss: 1.2306... Generator Loss: 4.8610
Epoch 1/1... Discriminator Loss: 1.0145... Generator Loss: 0.6993
Epoch 1/1... Discriminator Loss: 0.6839... Generator Loss: 1.2806
Epoch 1/1... Discriminator Loss: 0.6927... Generator Loss: 1.4810
Epoch 1/1... Discriminator Loss: 1.3397... Generator Loss: 0.4830

Epoch 1/1... Discriminator Loss: 1.6310... Generator Loss: 4.0758
Epoch 1/1... Discriminator Loss: 2.8799... Generator Loss: 0.0832
Epoch 1/1... Discriminator Loss: 2.0107... Generator Loss: 5.0411
Epoch 1/1... Discriminator Loss: 0.6215... Generator Loss: 1.3257
Epoch 1/1... Discriminator Loss: 1.2001... Generator Loss: 0.4919
Epoch 1/1... Discriminator Loss: 0.7601... Generator Loss: 2.7295
Epoch 1/1... Discriminator Loss: 0.9029... Generator Loss: 0.8618
Epoch 1/1... Discriminator Loss: 0.8372... Generator Loss: 1.2793
Epoch 1/1... Discriminator Loss: 1.0432... Generator Loss: 0.7578
Epoch 1/1... Discriminator Loss: 0.8425... Generator Loss: 1.7157
Epoch 1/1... Discriminator Loss: 2.1253... Generator Loss: 0.1876
Epoch 1/1... Discriminator Loss: 3.3150... Generator Loss: 4.8705
Epoch 1/1... Discriminator Loss: 1.4777... Generator Loss: 0.4093
Epoch 1/1... Discriminator Loss: 0.8708... Generator Loss: 0.9922
Epoch 1/1... Discriminator Loss: 0.6269... Generator Loss: 1.2198
Epoch 1/1... Discriminator Loss: 0.4698... Generator Loss: 1.8758
Epoch 1/1... Discriminator Loss: 0.6257... Generator Loss: 1.2384
Epoch 1/1... Discriminator Loss: 0.3764... Generator Loss: 1.9595
Epoch 1/1... Discriminator Loss: 0.6160... Generator Loss: 1.2343
Epoch 1/1... Discriminator Loss: 0.4199... Generator Loss: 2.5717

Epoch 1/1... Discriminator Loss: 0.7322... Generator Loss: 0.8661
Epoch 1/1... Discriminator Loss: 0.5194... Generator Loss: 2.3920
Epoch 1/1... Discriminator Loss: 1.1269... Generator Loss: 0.6545
Epoch 1/1... Discriminator Loss: 1.0955... Generator Loss: 2.4721
Epoch 1/1... Discriminator Loss: 1.8596... Generator Loss: 0.2519
Epoch 1/1... Discriminator Loss: 2.5279... Generator Loss: 5.4216
Epoch 1/1... Discriminator Loss: 0.8923... Generator Loss: 1.3665
Epoch 1/1... Discriminator Loss: 1.7757... Generator Loss: 0.7622
Epoch 1/1... Discriminator Loss: 2.4093... Generator Loss: 3.6284
Epoch 1/1... Discriminator Loss: 2.5979... Generator Loss: 0.1773
Epoch 1/1... Discriminator Loss: 1.8111... Generator Loss: 0.5613
Epoch 1/1... Discriminator Loss: 1.5251... Generator Loss: 3.1060
Epoch 1/1... Discriminator Loss: 1.5675... Generator Loss: 0.4256
Epoch 1/1... Discriminator Loss: 0.9175... Generator Loss: 2.0582
Epoch 1/1... Discriminator Loss: 1.2247... Generator Loss: 0.5254
Epoch 1/1... Discriminator Loss: 1.3152... Generator Loss: 2.0067
Epoch 1/1... Discriminator Loss: 1.4606... Generator Loss: 0.3773
Epoch 1/1... Discriminator Loss: 1.0203... Generator Loss: 2.5931
Epoch 1/1... Discriminator Loss: 1.1536... Generator Loss: 0.7789
Epoch 1/1... Discriminator Loss: 1.1052... Generator Loss: 0.6983

Epoch 1/1... Discriminator Loss: 1.0087... Generator Loss: 1.8723
Epoch 1/1... Discriminator Loss: 1.3513... Generator Loss: 0.4508
Epoch 1/1... Discriminator Loss: 0.9800... Generator Loss: 2.1883
Epoch 1/1... Discriminator Loss: 0.8975... Generator Loss: 0.9336
Epoch 1/1... Discriminator Loss: 0.9767... Generator Loss: 0.8024
Epoch 1/1... Discriminator Loss: 1.0086... Generator Loss: 0.9632
Epoch 1/1... Discriminator Loss: 1.3611... Generator Loss: 0.6824
Epoch 1/1... Discriminator Loss: 0.9967... Generator Loss: 0.9511
Epoch 1/1... Discriminator Loss: 0.9822... Generator Loss: 1.3644
Epoch 1/1... Discriminator Loss: 0.7350... Generator Loss: 1.4503
Epoch 1/1... Discriminator Loss: 1.0402... Generator Loss: 1.0415
Epoch 1/1... Discriminator Loss: 0.7830... Generator Loss: 1.6117
Epoch 1/1... Discriminator Loss: 1.3754... Generator Loss: 0.5075
Epoch 1/1... Discriminator Loss: 1.3899... Generator Loss: 3.3220
Epoch 1/1... Discriminator Loss: 1.3820... Generator Loss: 0.4647
Epoch 1/1... Discriminator Loss: 1.2320... Generator Loss: 1.0076
Epoch 1/1... Discriminator Loss: 2.2479... Generator Loss: 0.3507
Epoch 1/1... Discriminator Loss: 2.3489... Generator Loss: 3.0015
Epoch 1/1... Discriminator Loss: 3.6897... Generator Loss: 0.0896
Epoch 1/1... Discriminator Loss: 1.5516... Generator Loss: 3.3302

Epoch 1/1... Discriminator Loss: 1.5545... Generator Loss: 1.4004
Epoch 1/1... Discriminator Loss: 1.0422... Generator Loss: 0.9200
Epoch 1/1... Discriminator Loss: 1.1925... Generator Loss: 0.7289
Epoch 1/1... Discriminator Loss: 0.9783... Generator Loss: 0.9220
Epoch 1/1... Discriminator Loss: 1.0913... Generator Loss: 0.9278
Epoch 1/1... Discriminator Loss: 1.1042... Generator Loss: 1.4342
Epoch 1/1... Discriminator Loss: 1.4826... Generator Loss: 0.4160
Epoch 1/1... Discriminator Loss: 1.3160... Generator Loss: 1.6082
Epoch 1/1... Discriminator Loss: 1.5925... Generator Loss: 0.3906
Epoch 1/1... Discriminator Loss: 1.1722... Generator Loss: 1.8115
Epoch 1/1... Discriminator Loss: 1.1073... Generator Loss: 0.6887
Epoch 1/1... Discriminator Loss: 1.0619... Generator Loss: 0.8092
Epoch 1/1... Discriminator Loss: 1.3054... Generator Loss: 0.7173
Epoch 1/1... Discriminator Loss: 1.4136... Generator Loss: 0.8536
Epoch 1/1... Discriminator Loss: 1.8586... Generator Loss: 0.3303
Epoch 1/1... Discriminator Loss: 1.9618... Generator Loss: 2.3609
Epoch 1/1... Discriminator Loss: 1.7048... Generator Loss: 0.2878
Epoch 1/1... Discriminator Loss: 1.0754... Generator Loss: 1.3992
Epoch 1/1... Discriminator Loss: 1.0638... Generator Loss: 0.8265
Epoch 1/1... Discriminator Loss: 1.0891... Generator Loss: 1.5658

Epoch 1/1... Discriminator Loss: 2.1019... Generator Loss: 0.1873
Epoch 1/1... Discriminator Loss: 1.4727... Generator Loss: 2.0894
Epoch 1/1... Discriminator Loss: 1.2737... Generator Loss: 0.8595
Epoch 1/1... Discriminator Loss: 1.3341... Generator Loss: 0.5160
Epoch 1/1... Discriminator Loss: 1.2586... Generator Loss: 1.1166
Epoch 1/1... Discriminator Loss: 1.1419... Generator Loss: 0.7443
Epoch 1/1... Discriminator Loss: 1.2250... Generator Loss: 0.6462
Epoch 1/1... Discriminator Loss: 1.1826... Generator Loss: 0.8572
Epoch 1/1... Discriminator Loss: 1.3477... Generator Loss: 0.5733
Epoch 1/1... Discriminator Loss: 1.1710... Generator Loss: 0.6760
Epoch 1/1... Discriminator Loss: 1.2902... Generator Loss: 1.1621
Epoch 1/1... Discriminator Loss: 1.7367... Generator Loss: 0.3120
Epoch 1/1... Discriminator Loss: 1.6701... Generator Loss: 1.5297
Epoch 1/1... Discriminator Loss: 1.3476... Generator Loss: 0.5411
Epoch 1/1... Discriminator Loss: 1.1653... Generator Loss: 0.7997
Epoch 1/1... Discriminator Loss: 1.2789... Generator Loss: 0.7610
Epoch 1/1... Discriminator Loss: 1.0309... Generator Loss: 1.2468
Epoch 1/1... Discriminator Loss: 1.1525... Generator Loss: 0.7305
Epoch 1/1... Discriminator Loss: 1.0111... Generator Loss: 0.7537
Epoch 1/1... Discriminator Loss: 1.0844... Generator Loss: 1.3831

Epoch 1/1... Discriminator Loss: 1.6999... Generator Loss: 0.2983
Epoch 1/1... Discriminator Loss: 1.0692... Generator Loss: 1.2998
Epoch 1/1... Discriminator Loss: 1.2835... Generator Loss: 0.5948
Epoch 1/1... Discriminator Loss: 1.3365... Generator Loss: 0.8625
Epoch 1/1... Discriminator Loss: 1.3652... Generator Loss: 0.6666
Epoch 1/1... Discriminator Loss: 1.2807... Generator Loss: 1.4351
Epoch 1/1... Discriminator Loss: 1.8497... Generator Loss: 0.2543
Epoch 1/1... Discriminator Loss: 1.8441... Generator Loss: 1.7496
Epoch 1/1... Discriminator Loss: 1.6588... Generator Loss: 0.3408
Epoch 1/1... Discriminator Loss: 1.3582... Generator Loss: 0.7058
Epoch 1/1... Discriminator Loss: 1.3382... Generator Loss: 1.1245
Epoch 1/1... Discriminator Loss: 1.4074... Generator Loss: 0.5893
Epoch 1/1... Discriminator Loss: 1.5509... Generator Loss: 0.9778
Epoch 1/1... Discriminator Loss: 1.5013... Generator Loss: 0.3653
Epoch 1/1... Discriminator Loss: 1.4238... Generator Loss: 1.2062
Epoch 1/1... Discriminator Loss: 1.4703... Generator Loss: 0.4496
Epoch 1/1... Discriminator Loss: 1.4199... Generator Loss: 0.9667
Epoch 1/1... Discriminator Loss: 1.7203... Generator Loss: 0.5214
Epoch 1/1... Discriminator Loss: 1.4920... Generator Loss: 0.7958
Epoch 1/1... Discriminator Loss: 1.4521... Generator Loss: 0.6160

Epoch 1/1... Discriminator Loss: 1.4030... Generator Loss: 0.7247
Epoch 1/1... Discriminator Loss: 1.4111... Generator Loss: 0.5219
Epoch 1/1... Discriminator Loss: 1.3363... Generator Loss: 0.9131
Epoch 1/1... Discriminator Loss: 1.6512... Generator Loss: 0.3845
Epoch 1/1... Discriminator Loss: 1.5477... Generator Loss: 0.6687
Epoch 1/1... Discriminator Loss: 1.4557... Generator Loss: 0.9296
Epoch 1/1... Discriminator Loss: 1.2556... Generator Loss: 0.7335
Epoch 1/1... Discriminator Loss: 1.4006... Generator Loss: 0.5405
Epoch 1/1... Discriminator Loss: 1.2896... Generator Loss: 0.8197
Epoch 1/1... Discriminator Loss: 1.3254... Generator Loss: 1.1323
Epoch 1/1... Discriminator Loss: 1.3896... Generator Loss: 0.4067
Epoch 1/1... Discriminator Loss: 1.1970... Generator Loss: 1.5438
Epoch 1/1... Discriminator Loss: 1.1564... Generator Loss: 0.5792
Epoch 1/1... Discriminator Loss: 1.0726... Generator Loss: 0.7881
Epoch 1/1... Discriminator Loss: 1.1862... Generator Loss: 1.2657
Epoch 1/1... Discriminator Loss: 1.8071... Generator Loss: 0.2465
Epoch 1/1... Discriminator Loss: 3.5741... Generator Loss: 3.2860
Epoch 1/1... Discriminator Loss: 1.2074... Generator Loss: 0.9677
Epoch 1/1... Discriminator Loss: 1.7867... Generator Loss: 0.2955
Epoch 1/1... Discriminator Loss: 1.3387... Generator Loss: 0.6066

Epoch 1/1... Discriminator Loss: 1.4770... Generator Loss: 1.0318
Epoch 1/1... Discriminator Loss: 1.6811... Generator Loss: 0.4618
Epoch 1/1... Discriminator Loss: 1.6447... Generator Loss: 0.6318
Epoch 1/1... Discriminator Loss: 1.5252... Generator Loss: 1.0820
Epoch 1/1... Discriminator Loss: 1.5466... Generator Loss: 0.6015
Epoch 1/1... Discriminator Loss: 1.4521... Generator Loss: 0.6514
Epoch 1/1... Discriminator Loss: 1.2714... Generator Loss: 1.1834
Epoch 1/1... Discriminator Loss: 1.3987... Generator Loss: 0.5497
Epoch 1/1... Discriminator Loss: 1.5780... Generator Loss: 0.5558
Epoch 1/1... Discriminator Loss: 1.5332... Generator Loss: 0.6049
Epoch 1/1... Discriminator Loss: 1.3073... Generator Loss: 0.8751
Epoch 1/1... Discriminator Loss: 1.3147... Generator Loss: 0.6686
Epoch 1/1... Discriminator Loss: 1.1929... Generator Loss: 0.9972
Epoch 1/1... Discriminator Loss: 1.0877... Generator Loss: 0.6947
Epoch 1/1... Discriminator Loss: 0.9983... Generator Loss: 1.0626
Epoch 1/1... Discriminator Loss: 1.1854... Generator Loss: 0.6055
Epoch 1/1... Discriminator Loss: 1.2539... Generator Loss: 0.9563
Epoch 1/1... Discriminator Loss: 1.7215... Generator Loss: 0.3276
Epoch 1/1... Discriminator Loss: 1.7158... Generator Loss: 1.7582
Epoch 1/1... Discriminator Loss: 1.4181... Generator Loss: 0.4484

Submitting This Project¶

When submitting this project, make sure to run all the cells before saving the notebook. Save the notebook file as "dlnd_face_generation.ipynb" and save it as a HTML file under "File" -> "Download as". Include the "helper.py" and "problem_unittests.py" files in your submission.